433013 Deployment of Explicit Model Predictive Controller Onboard a Mini Fuel Cell Vehicle

Wednesday, November 11, 2015: 8:30 AM
250B (Salt Palace Convention Center)
Amit M. Manthanwar, Imperial College London, London, United Kingdom and Efstratios N. Pistikopoulos, Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX

Deployment of Explicit Model Predictive Controller Onboard a Mini Fuel Cell Vehicle

Amit M. Manthanwar and Efstratios N. Pistikopoulos

Department of Chemical Engineering, Imperial College London, England, UK

Artie McFerrin Department of Chemical Engineering, Texas A&M University, USA


Targeted Session: 07000 Recent Advances in Fuel Cell and Battery Technologies

Fuel cells are electrochemical devices that convert hydrogen or hydrogen rich fuel directly into electricity. Fuel cells are more efficient compared to other combustion based technologies and offer a wide range of tangible environmental, economic and operational benefits. However, control is critical to the robust operation of the system—ultimately essential for maintaining uninterruptable power demand under load variations, [1]. Furthermore, fuel cell system integration to automotive powertrain brings additional challenges affecting durability and performance of the overall system. The majority of fuel cell system failures and forced outages (~90%) are due to lack of system integration, [2].

The real time control of fuel cell vehicle using a well established technology of model predictive control requires an accurate mathematical model of the system. The dynamic phenomenon taking place inside fuel cell involves complex interactions of mass transport, energy transport and electrochemical kinetics. This complexity further increases when the fuel cell dynamics are coupled with the vehicle transmission dynamics. These uncertain dynamics determine the overall operational performance of the vehicle. In addition, the real time control requires computing machinery to run the optimisation algorithms. This compounds the manufacturing and operational costs. In order to resolve these issues, we present (a) an experimentally validated dynamic model of the polymer electrolyte membrane fuel cell; (b) an experimentally validated model of DC motors used in a vehicle powertrain; and (c) generic control framework with deployment strategy of explicit model predictive controller, [3], for the efficient operation of the fuel cell vehicle propulsion system.

This work presents a prototype fuel cell vehicle with step by step procedure to deploy explicit MPC on a fast ‘mbed’ platform based on a high performance ARMr CortexTM M3 microcontroller. Our analysis of the experimentally achieved driving cycle results  demonstrate that significant reduction in operational costs can be achieved while simultaneously improving the overall performance the fuel cell vehicle. Such modelling and control framework is critical to effective operation as well as preventive maintenance of the fuel cell vehicle. Thus, we conclude that proposed “MPC-on-a-chip” strategy is the best available option for making smarter decisions in operating fuel cell vehicles reliability and further improving their competitive advantage.


[1] C. Ziogou, E. N. Pistikopoulos, M. C. Georgiadis, S. Voutetakis, and S. Papadopoulou, “Empowering the performance of advanced nmpc by multiparametric programming–an application to a pem fuel cell system,” Ind. Eng. Chem. Res., vol. 52, pp. 4863–4873, Mar. 2013.

[2] K. Wipke, S. Sprik, J. Kurtz, T. Ramsden, C. Ainscough, and G. Saur, “Final results from u.s. fcev learning demonstration,” in EVS26 International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, (Los Angeles, California), May 6–9 2012.

[3] E. N. Pistikopoulos, “From multi-parametric programming theory to mpc-on-a-chip multi-scale systems applications,” Computers and Chemical Engineering, no. 0, 2012.

Extended Abstract: File Not Uploaded
See more of this Session: Recent Advances in Fuel Cell and Battery Technologies
See more of this Group/Topical: Transport and Energy Processes